Posts
http://www.extremetech.com/extreme/105067-mit-creates-brain-chip
By Sebastian Anthony
With 400 transistors and standard CMOS manufacturing techniques, a group of MIT researchers have created the first computer chip that mimics the analog, ion-based communication in a synapse between two neurons. This is the bleeding edge of brain-like (neural network) processors, but let’s put this into perspective: The human brain has around 100 billion neurons, and each neuron can be connected to thousands of others with synapses. There are trillions or quadrillions of synapses in the human brain.
There are three ways in which you can attempt to model the human brain in silicon: You can throw more and more processing power at the problem until you reach brain-like capabilities; you can make a learning crossbar switch that simulates the multiple connections between neurons (but this gets very big very quickly); or you can go the whole hog and use analog technology to actually mimic the chemical, ion-based communication channels that flow between synapses. It is this last category which the new chip from MIT falls into.
Contrary to the highschool (and House) education that you may have received, the neurons in animal brains, including the fantastic sample ensconced by your skull, do not simply “fire.” Every neuron has thousands of synapses, and the flow of electricity over those synapses is controlled by the flow of ions; charged molecules of sodium, potassium, chloride, and calcium. It is the concentration of these ions, the timing of the electrical pulses generated by the neuron, and myriad other factors, that ultimately govern the massively-parallel computational power of a brain.
Lt Commander Data
Scientists and engineers have tried to fashion brain-like neural networks before, of course, but transistor-transistor logic is fundamentally digital — and the brain is completely analog. Neurons do not suddenly flip from “0″ to “1″ — they can occupy an almost-infinite scale of analog, in-between values. You can approximate the analog function of synapses by using fuzzy logic (and by ladling on more processors), but that approach only goes so far.
MIT’s chip — all 400 transistors (pictured below) — is dedicated to modeling every biological caveat in a single synapse. “We now have a way to capture each and every ionic process that’s going on in a neuron,” says Chi-Sang Poon, an MIT researcher who worked on the project. The next step? Scaling up the number of synapses and building specific parts of the brain, such as our visual processing or motor control systems. The long-term goal would be to provide bionic components that augment or replace parts of the human physiology, perhaps in blind or crippled people.
Of course, with truly analog processors comes real artificial intelligence, too — and not the kind that requires megawatts of power and a hangar full of server racks. With current state-of-the-art technology it takes hours or days to simulate a simple brain circuit. With MIT’s brain chip, the simulation is “faster than the biological system itself.” Gulp.
